Redpanda 25.2: Advancing Apache Iceberg™ integration for the real-time lakehouse

With the release of Redpanda 25.2, we’re advancing our investment in Iceberg capabilities to deliver on the promise of a real-time open lakehouse architecture. This update reflects our continued commitment to enabling enterprises to unify streaming and analytics workloads with simplicity, efficiency, and architectural flexibility. 

From integrations with Databricks Unity Catalog and AWS Glue Catalog, to expanded schema support for Iceberg Topics — Redpanda 25.2 elevates Iceberg to a first-class capability in the Redpanda ecosystem.

Additionally, this release introduces features like Schema Registry Authorization key updates designed to enhance data lake integration, simplify schema management, strengthen security protocols, and broaden our Apache Kafka^® client compatibility for your event streaming workloads.

Catalog Integration: AWS Glue, Databricks Unity, and beyond

Redpanda 25.2 expands your options for Iceberg catalogs. Whether you’re on AWS using Glue Data Catalog, a Databricks customer using Unity Catalog, a Snowflake pro using Open Catalog, or exploring options like Apache Polaris, Redpanda provides a way to deliver real-time insights, low-latency ingestion, and open interoperability across systems.

We’re focused on giving you open, flexible integrations without vendor lock-in or managed service traps. Your buckets, your catalogs, your control.

AWS Glue Data Catalog Support for Iceberg Topics: Streamlining data lake integration 

Redpanda 25.2 now supports using AWS Glue as the Iceberg REST Catalog for Redpanda Iceberg Topics. This means if you're using Glue Catalog, you can effortlessly integrate it with your Redpanda Iceberg Topics, making it extra simple to connect AWS tools with streaming data.

This is a game-changer if you don't have a separate lakehouse or catalog solution. It's about bringing your real-time streaming data directly into your existing AWS data ecosystem, simplifying your architecture, and giving you instant access to fresh data for analytics and AI initiatives. It's like finding that missing piece of the puzzle, and it fits perfectly! 

For more on using Iceberg catalogs with Redpanda, see our docs. 

Databricks + Redpanda: Building the real-time Iceberg lakehouse

Speaking of Databricks and Redpanda, we believe your streaming data should flow straight into your lakehouse — in real time, with full control. That’s why Redpanda Iceberg Topics integrate seamlessly with open catalog solutions like Databricks Unity Catalog.

Together, Redpanda and Databricks make a powerful pair. Here's what you can do with it:

• Your real-time event streams from Redpanda become queryable in Databricks using Spark or DB SQL as Managed Iceberg tables.
• You can feed your AI/ML models, dashboards, and batch pipelines fresh streaming data without waiting on batch ETL jobs.
• You get to keep full control of your infrastructure with a self-managed, bring-your-own-cloud (BYOC) approach.

If you’re using Databricks as your primary analytics and AI platform, Redpanda slots in as the real-time streaming engine, bringing speed, simplicity, and openness to your lakehouse architecture. It's the best of both worlds: real-time insights for operational use cases and a robust, scalable lakehouse for enterprise analytics and AI.

For a deeper look, check out our joint write-up: The Real-Time Open Lakehouse with Redpanda and Databricks.

Iceberg support for JSON schemas: Expanding schema flexibility 

You know that friend who always orders the same thing at a restaurant? Sometimes, it's just about comfort and familiarity. For many of our customers, JSON is that comfort food for their data. We've always supported Protobuf and Avro for Iceberg schemas, but we heard you, so with Redpanda 25.2, we're bringing Iceberg support for JSON Schemas! 

That means if you're already using JSON as your data format for Kafka workloads and have your JSON schemas neatly organized in the schema registry, you can now seamlessly map them to Iceberg schemas for your Iceberg Topics. 

This update is crucial for customers who rely on JSON for their existing Kafka setups. It means less re-tooling and more doing, allowing you to leverage your current investments and accelerate your adoption of Iceberg Topics for building efficient, real-time data lakes. Your data, your format.

Schema Registry Authorization: Enhancing security and governance 

Redpanda 25.2 introduces Schema Registry Authorization, providing a fine-grained authorizer for Redpanda's Schema Registry that offers capabilities similar to Confluent's ACL Authorizer.

This means you can now set explicit access control lists (ACLs) on specific subjects and schemas within our Redpanda CLI (rpk), giving you granular control over who can do what with your precious schema data. We're talking about giving you the keys to the castle, but only to the rooms you want to let people into. 

Why explicit ACLs? Turns out that relying on implicit ACLs based on topic names (like Confluent's Topic ACL Authorizer) just isn't cutting it for complex enterprise environments, and even Confluent isn't recommending it anymore (no support in KRaft mode). So, we want to make sure your schemas are locked down as tightly as possible to protect workloads from destructive or incompatible changes to schemas by other users.

Users will be able to set privileges (ACLs) on specific Subjects and Schemas as well as the entire registry. We’ll also support the assignment of these subject/schema ACLs via Redpanda RBAC (roles) and assign ACLs directly to users. Management of these ACLs will be available via the Schema Registry API endpoint, and via new rpk commands within rpk security acl. 

Redpanda also sports enhanced audit logs for schema registry. Audit logs now record both the sensitive ACL manipulation actions taken by admins and provide a full audit trail of schema registry access, reporting every attempt to access a registry resource. This makes it easy to answer the “who, what, when, and why” of access, and even the specific ACLs behind the decision to grant or deny each request. 

With Schema Registry Authorization, Redpanda 25.2 bolsters data governance and compliance, prevents mistakes that can interrupt workloads, and secures sensitive schema information within the enterprise.

New client support: Your npm install just got better with kafka-javascript!

For our developer friends building modern backend applications, APIs, and real-time services, Node.js is often the go-to language. However, the ecosystem of Kafka clients for Node.js has, historically, been a bit fragmented.

That's why we're thrilled to announce that as of the 25.2 release, Redpanda officially certifies kafka-javascript — and we promise you won't need to await clarity on its performance or stability. This is the new, mainstream Node.js client for Kafka, built and sponsored by Confluent (npm, GitHub), that we have tested and verified to work with Redpanda.

So go forth and build with confidence!

Check out our full list of supported clients for more details.

Keep those tables flowing 👋

When it comes to Apache Iceberg, we believe that your tables should be fully open — in buckets that you own, not trapped in managed service and upcharged for object storage you don't control. Redpanda delivers on Iceberg's open promise, with BYOC and self-managed enterprise streaming options that adapt to your use case.

That's a wrap on the Redpanda 25.2 release highlights! We're constantly working to make Redpanda the simple, fast, and secure streaming platform for all your real-time and AI workloads. Because at Redpanda, we believe your data infrastructure should be a joy, not a chore. 

Want to see 25.2 in action? Join us for our upcoming Tech Talk with Databricks on August 20th. We’ll dive deeper into building a real-time architecture with Redpanda Iceberg Topics and Databricks Unity Catalog. Register here to save your spot.

Stay tuned for more updates, and as always, happy streaming!